Is eyewitness testimony reliable?

As most crimes include some sort of violence, this can cause eyewitness unreliability. Clifford and Hollin (1981) concluded that violence has a negative recall on memory, despite most of us believing that violence actually improves recall. Clifford and Hollin showed participants video footage of a man either stealing a woman’s handbag and roughly pushing her against a wall or asking her for directions. Recall was less accurate for the violent footage, this might have been because violence causes interferences in memory and therefore, does not allow rehearsal to take place/information to be transferred into the LTM.

The use of weapons and violence cause poor recall as peripheral details of events aren’t recalled. An investigation was carried out by Pickel (1988) to test whether the presence of a weapon affected the ability to recall. Participants were showed a 2 minute video clip of a man walking up to a hairdresser’s receptionist with either nothing in his hand, a pair of scissors, a gun, a wallet or a raw chicken. Participants where then asked what the man was doing and to recognise him from a line-up. The results showed that the unseal objects had the most deleterious effect on recall but not on line-up identification. The usualness of a weapon in a hairdressers’ could explain why memory is impaired, rather than the sense of threat/danger it imposes.

Other factors that affect the accuracy of eye witness testimony are time distortion. Loftus et all (1987) found that the estimated duration of an event is distorted. He found that the estimated duration of the footage shown to participant was five times greater than it actually was.  

Elizabeth Loftus researched the influence of post-event information and found out that it is incorporated into the original memory when recalled.  

However, eye witness testimony can also be reliable as witnesses believe that the interviewer has no knowledge of the event and will less likely to try and ‘please’ the interviewer by giving them the answer they believe is right. Yuille and Cutshall’s study of statements given four months after witnessing a gun shooting found that witness recall was very accurate and not affected by the use of leading questions. Also, Thomson et al. (1997) studied statements given by the survivors of the sunken river boat Marchioness and found that, despite extreme emotional trauma, recall was very accurate even after many months.

Eyewitness testimony could be improved by the use of cue-dependency. A cue is something that is present during the learning the learning phase (the event of the crime) which, reinstated when recalling, could improve as it act as a trigger for the original memory.  For example, witnesses could revisit the scene or there could be a reconstruction of the event and this could help ‘trigger’ some more memories/details of the event. However, the reconstruction could cause retroactive interference as memories of the reconstruction event could interfere with memories of the actual event. Also, false memories may occur as it would be hard to reconstruct the event accurately.  One way of avoiding of the first problem could be to the questioning at the scene which could be recorded.

Nevertheless, eyewitness testimony is vital as dismissing it altogether would render some crime, such as rape, untriable.